picasso: Pathwise Calibrated Sparse Shooting Algorithm

Computationally efficient tools for fitting generalized linear model with convex or non-convex penalty. Users can enjoy the superior statistical property of non-convex penalty such as SCAD and MCP which has significantly less estimation error and overfitting compared to convex penalty such as lasso and ridge. Computation is handled by multi-stage convex relaxation and the PathwIse CAlibrated Sparse Shooting algOrithm (PICASSO) which exploits warm start initialization, active set updating, and strong rule for coordinate preselection to boost computation, and attains a linear convergence to a unique sparse local optimum with optimal statistical properties. The computation is memory-optimized using the sparse matrix output.

Install the latest version of this package by entering the following in R:
install.packages("picasso")
AuthorJason Ge, Xingguo Li, Mengdi Wang, Tong Zhang, Han Liu and Tuo Zhao
Date of publication2016-10-06 20:14:19
MaintainerJason Ge <jiange@princeton.edu>
LicenseGPL-2
Version0.5.4

View on CRAN

Files

inst
inst/doc
inst/doc/vignette.Rnw
inst/doc/PICASSO.pdf
inst/doc/vignette.pdf
src
src/IRLS_solver.c
src/mathutils.c
src/mathutils.h
src/picasso_gaussian_cov.c
src/picasso_poisson.c
src/IRLS_solver.h
src/picasso_gaussian_naive.c
src/picasso_logit.c
NAMESPACE
data
data/eyedata.rda
data/datalist
R
R/picasso.R R/poisson_solver.R R/picasso.poisson.R R/logit_solver.R R/picasso.gaussian.R R/picasso.logit.R R/gaussian_solver.R
vignettes
vignettes/vignette.Rnw
vignettes/PICASSO.pdf
MD5
build
build/vignette.rds
DESCRIPTION
man
man/print.poisson.Rd man/coef.logit.Rd man/picasso.Rd man/coef.poisson.Rd man/print.gaussian.Rd man/coef.gaussian.Rd man/eyedata.Rd man/predict.poisson.Rd man/predict.gaussian.Rd man/print.logit.Rd man/predict.logit.Rd man/plot.logit.Rd man/plot.gaussian.Rd man/plot.poisson.Rd man/picasso-package.Rd

Questions? Problems? Suggestions? or email at ian@mutexlabs.com.

Please suggest features or report bugs with the GitHub issue tracker.

All documentation is copyright its authors; we didn't write any of that.